EP0950197B1 - Device and method for nuclear locating by iterative computing of barycenter, and application to gamma-cameras - Google Patents

Description

Technical area

The present invention relates to a device for determination of the position of an inducing event a signal in photodetectors, this position being, for example, spotted with respect to the set photodetecers. Such a position can be spotted by the center of gravity of the event in a landmark linked to photodetectors.

The invention applies in particular to the determining the position of an event from signals provided by photomultipliers equipping a gamma camera, the position being located relative to to the photomultipliers themselves. We hear by gamma camera a camera sensitive to gamma radiation (Γ). Such cameras are used in particular for medical imaging purposes.

State of the art

At present, most gamma cameras used in nuclear medicine are cameras operating on the principle of the cameras of Anger type. We can refer to this subject in the document U.S. 3,011,057.

In particular, gamma cameras allow visualize the distribution, in an organ, of molecules marked by a radioactive isotope injected to the patient.

The structure and operation of a gamma camera are described and summarized below in reference to Figures 1, 2A and 2B appended.

Figure 1 shows a detection head 10 a gamma camera arranged in front of an organ 12 containing molecules labeled with a radioactive isotope.

The detection head 10 comprises a collimator 20, a scintillator crystal 22, a light guide 24 and a plurality of photomultiplier tubes 26 juxtaposed so as to cover one side of the guide of 24. The scintillator is, for example, a NaI crystal (Tℓ).

The purpose of the collimator 20 is to select from all gamma radiation 30 emitted by the organ 12 those who reach the head of sensing substantially under normal incidence. The selective nature of the collimator makes it possible to increase resolution and sharpness of the image produced. However, the increase in the resolution is done at detriment of sensitivity. For example, for about 10,000 γ photons emitted by the organ 12, a single photon is actually detected.

The γ photons having crossed the collimator reach the scintillator crystal 22 where almost each photon γ is converted into a plurality of bright photons. In the rest of the text we refer to event each interaction of a gamma photon with the crystal, causing scintillation.

Photomultipliers 26 are designed to send a proportional electrical pulse number of light photons received from the scintillator for each event.

For a scintillation event to to be located more precisely, the photomultipliers 26 are not directly contiguous to the scintillator crystal 22 but are separated from this last by the light guide 24.

Photomultipliers emit a signal whose amplitude is proportional to the quantity total light produced in the scintillator by a gamma radiation, that is to say, proportional to its energy. However, the individual signal of each photomultiplier also depends on the distance that the separates from the interaction point 30 of gamma radiation with the scintillator material. Indeed, every photomultiplier delivers a current pulse proportional to the luminous flux he has received. In the example of Figure 1, small graphs A, B, C show that photomultipliers 26a, 26b and 26c located at different distances from an interaction point 30 deliver signals with amplitudes different.

The position of the interaction point 30 of a gamma photon is calculated in the gamma-camera from signals from all of the photomultipliers by weighting barycentric contributions of each photomultiplier.

The principle of barycentric weighting such that it is implemented in Anger-type cameras appears more clearly with reference to FIGS. 2A and 2B appended.

Figure 2A shows the electrical wiring of a detection head 10 of a gamma camera, which connects this camera to a unit forming an image. The detection head has a plurality of photomultipliers 26.

As shown in FIG. 2B, each photomultiplier 26 of the detection head is associated with four resistors denoted RX ^- , RX ⁺ , RY ^- and RY ⁺ . The values of these resistors are specific to each photomultiplier and depend on the position of the photomultiplier in the detection head 10.

The resistors RX ^- , RX ⁺ , RY ^- and RY ⁺ of each photomultiplier are connected to the output 50 of said photomultiplier, shown in FIG. 2B with a current generator symbol. They are, on the other hand, respectively connected to common collecting lines denoted LX ^- , LX ⁺ , LY ^- , LY ⁺ , in FIG. 2A.

The lines LX ^- , LX ⁺ , LY ^- and LY ⁺ are in turn connected respectively to analog integrators 52X ^- , 52X ⁺ , 52Y ^- , 52Y ⁺ , and via these to analog / digital converters. 54X ^- , 54X ⁺ , 54Y ^- , 54Y ⁺ . The output of the converters 54X ^- , 54X ⁺ , 54Y ^- , 54Y ⁺ is directed to a digital operator 56. The lines LX ^- , LX ⁺ , LY ^- , LY ⁺ are also connected to a common channel, called energy path. This channel also comprises an integrator 57 and an analog / digital converter 58 and its output is also directed towards the operator 56.

With the device of FIG. 2, the position of the interaction is calculated according to the following equations (US-4,672,542): X = X + -X - X + + X - and Y = Y + -Y - Y + + Y - where X and Y indicate the coordinates in two orthogonal directions of the position of the interaction on the crystal and in which X ⁺ , X ^- , y ⁺ , Y ^- respectively indicate the weighted signals delivered by the integrators 52X ⁺ , 52X ^- , 52Y ⁺ , 52Y ^- .

The values of X and Y as well as the energy total E of the gamma ray that interacted with the crystal are established by the digital operator 56. These values are then used for constructing an image as described, for example, in the document FR-2,669,439.

The calculation of the position of the interaction is tainted with uncertainty about fluctuations Poisson statistics of the number of light photons and the number of photoelectrons produced for each event, that is to say for each gamma photon detected. The standard deviation of fluctuation is all the lower than the number of photons or photoelectron is high. Because of this phenomenon, he should collect the light most carefully possible. The intrinsic spatial resolution of the camera is characterized by the width at half height of the distribution of calculated positions for the same collimated point source placed on the crystal scintillator.

For gamma rays with an energy of 140 keV, the resolution is generally of the order of 3 to 4 mm.

The energy of a detected gamma photon is calculated by summing the contributions of all photomultipliers having received light. She is also tainted by a statistical fluctuation. The energy resolution of the camera is characterized by the ratio of the halfway width of the distribution of calculated energies to the average value distribution, for the same source.

The energy resolution is usually the order of 9 to 11% for gamma rays of an energy 140 keV.

Finally, a gamma-camera type Anger has the advantage of being able to calculate time real the center of gravity of the photomultiplier signals with very simple means.

Indeed, the system described previously has a limited number of components. Moreover, the resistors used to inject the signal from photomultipliers in the collector lines are very inexpensive.

One-time improvements to the camera ANGER type have been proposed. U.S. Patent No. 5,576,547 proposes a method to correct the calculation of the total energy received by the detectors and deduce a corrected position of the event.

Correction tables are established in building for known event positions of histograms of the energy captured by the different detectors.

These tables are then used to correct the position of the event.

A camera such as the ANGER camera presents however also a major disadvantage which is a rate counting reduced. Count rate means the number of events, that is, interactions between a γ photon and the scintillator, that the camera is able to process per unit of time.

One of the limitations of the count rate comes from the fact that the camera is incapable to deal with two events occurring substantially simultaneously at distinct points of the crystal scintillator.

In fact, simultaneous but geometrically distinct events give rise to electrical signals which accumulate in the collector lines LX ^- , LX ⁺ , LY ^- and LY ⁺ and which can no longer be distinguished. These events are also "lost" for the formation of an image.

This concern of taking into account and correction stacks is present in the patent application European Commission No 0 252 566.

The device described in this application includes means for digitizing the signal provided by each detector and the energy of each photodetector is calculated with a correction for to keep track of the energy provided by events occurring while the current event is still In progress.

Limitation of the count rate is not too much stress in the techniques traditional medical imaging. Indeed, as indicated above, the collimator stops a very large number of gamma rays and only a small number events are actually detected.

Gamma cameras are however used also in two other imaging techniques medical system where the limitation of the count rate is a crippling constraint.

These techniques are the so-called "attenuation correction by transmission" and "PET (Position Emission Tomography) Coincidentally ".

The mitigation correction technique by transmission is to take into account, when formation of a medical image, of the own attenuation tissue of the patient surrounding the examined organ. For know this attenuation, we measure the transmission gamma radiation to a gamma camera through the body of the patient. For this purpose we take place to the patient between a very active external source and the detection head of the gamma camera. So, during the measurement of transmitted radiation, a high number events take place in the scintillator crystal. The high number of events per unit of time increases also the probability of having multiple events substantially simultaneous. Anger-type camera classic then turns out to be inappropriate.

The PET technique consists of injecting the patient with an element such as F ¹⁸ capable of emitting positrons. The annihilation of a positron and an electron releases two γ photons emitted in opposite directions and having an energy of 511 keV. This physical phenomenon is used in the PET imaging technique. In this technique a gamma camera is used with at least two detection heads arranged on either side of the patient. The detection heads used are not equipped with a collimator. Indeed, an electronic processing of the information, said treatment of coincidence, makes it possible to select among the events those which coincide temporally, and to calculate thus the trajectory of the gamma photons.

The detection heads are therefore subject to high gamma radiation flux. Gamma cameras Anger classics have a count rate generally too limited for such an application.

As an indication, an Anger type gamma camera can function normally with a detection of 1.10 ⁵ events per second, while in PET imaging it takes at least 1.10 ⁶ events per second for normal operation.

Another limitation of gamma-cameras of the Anger type, described above, is that the calculation of the centroid of an event is fixed permanently by the construction of the detection head and in particular by the choice of resistors RX ^- , RX ⁺ , RY ^- , RY ⁺ for each photomultiplier. Similarly, the calculation of the energy is fixed by the wiring of photomultipliers on a common path (energy path).

It is therefore necessary to develop devices and methods for the use of gamma cameras with high counting rates.

In addition, we wish to develop cameras for which the determination of the center of gravity, or location, of an event, presents good linearity and resolution characteristics Space.

Presentation of the invention

a) une étape de numérisation du signal délivré par chaque photodétecteur, et de calcul d'une valeur N_i,j représentative de l'énergie du signal délivré par chaque photodétecteur,

b) calcul, en fonction des N_ij et de la position des photodétecteurs, d'une position brute P₀ de l'événement, par rapport à l'ensemble des photodétecteurs,

c) détermination de la distance d_i,j de chacun des photodétecteurs par rapport à la position P₀,

d) calcul d'une valeur corrigée N'_i,j=F(N_ij), où F est une fonction :

• qui réduit (respectivement : ne modifie pas) la valeur N_ij pour le photodétecteur correspondant à P₀ et pour un certain nombre N₁ de photodétecteurs autour de P₀ (par exemple : les N₁ photodétecteurs de première couronne),
• qui accroít, ou qui ne modifie pas (respectivement : accroít) la valeur N_ij pour un certain nombre de photodétecteurs N₂ situés autour des N₁ photodétecteurs précédents (par exemple : les N₂ photodétecteurs de deuxième couronne si les N₁ photodétecteurs. sont ceux de première couronne ; et les photodétecteurs de première couronne, ou de première et de deuxième couronnes si N₁=0),
• qui tend vers 0 au-delà,

e) calcul d'une nouvelle position P₁ de l'événement, en fonction de la position des photodétecteurs et des Ni_,j.

The subject of the invention is a method for determining the position P ₀ of an event with respect to a set of N photodetectors, this event inducing a signal in the N photodetectors, this method comprising the following steps:

a) a step of digitizing the signal delivered by each photodetector, and calculating a value N _{i, j} representative of the energy of the signal delivered by each photodetector,

b) calculating, as a function of the N _ij and the position of the photodetectors, a raw position P ₀ of the event, with respect to the set of photodetectors,

c) determining the distance d _{i, j} of each of the photodetectors with respect to the position P ₀ ,

d) calculating a corrected value N ' _{i, j} = F (N _ij ), where F is a function:

• which reduces (does not modify) the value N _ij for the photodetector corresponding to P ₀ and for a certain number N ₁ of photodetectors around P ₀ (for example: the N ₁ photodetectors of the first ring),
• which increases, or does not modify (respectively: increases) the value N _ij for a certain number of photodetectors N ₂ situated around the N ₁ previous photodetectors (for example: the N ₂ photodetectors of the second ring if the N ₁ photodetectors are those of first crown, and the photodetectors of first crown, or first and second crowns if N ₁ = 0),
• which tends to 0 beyond,

e) calculating a new position P ₁ of the event, as a function of the position of the photodetectors and Ni _{, j} .

Such a method makes it possible to process digitized data, and makes it possible to produce a position signal P ₁ of the event with respect to all the N photodetectors.

The iterative nature of the process according to the invention gives it a good spatial resolution and good linearity.

The calculation of the raw position P ₀ of the event can be performed by calculating the barycentric coordinates (X ₀ , Y ₀ ) of said event as a function of the N _{i, j} and the position (Xc _{i, j} , Yc _{i, j} ) of each photodetector.

Similarly, the calculation of the position P ₁ of the event can be performed by calculating the barycentric coordinates (X ₁ , Y ₁ ) of said event as a function of the N ' _{i, j} and the position (XC _{i, j} , YC _{i, j} ) of each photodetector.

The method according to the invention, because of the choice of the function F, reduces the contribution of the photodetector which is in front of the presumed position of the event (and, possibly, of a number N ₁ of photodetectors around this one): indeed, this (these) photodetector (s) bring (s) little information on the value of the position, or the barycentre of the event. Moreover, this function F gives an increased preponderance to the photodetectors located beyond that corresponding to the raw position P ₀ of the event and the N ₁ photodetectors.

One can choose, for example: N ₁ = 0 or N ₁ = 8 (first ring, especially for a square distribution of photodetectors).

A particular treatment can be realized for the photodetectors which are situated at the edge of the field of N photodetectors. This processing consists of the following additional step: modifying the value of the position (XC _{i, j} , YC _{i, j} ) to a new value (XC ' _{i, j} , YC' _{i, j} ), one at minus values | X '| and | Y '| being greater than | X | and | Y |. Thus, the field edge photodetectors have their modified weight, but also their position: the method according to the invention modifies both their contribution and their weight in the calculation of the position or the center of the event: this allows a enlargement of the field and improvement of the linearity.

Center of gravity calculations can be made sequentially. The treatment time is then all the longer as the number of Photodetectors involved in the calculation is great.

b₁) une sous-étape de détermination, pour chaque colonne i :

• de la contribution de la colonne à l'énergie totale induite par l'événement dans l'ensemble des photodétecteurs,
• de la contribution de la colonne au barycentre, en X, de l'événement,
• de la contribution de la colonne au barycentre, en Y, de l'événement.

b₂) une sous-étape de détermination :

• de l'énergie totale induite par l'événement dans l'ensemble des photodétecteurs,
• des coordonnées barycentriques (X₀, Y₀) de l'événement par rapport aux N photodétecteurs.

Alternatively, parallel data processing can be performed. Thus, the determination of the barycentric coordinates (X ₀ , Y ₀ ) of the event may comprise the following sub-steps:

b ₁ ) a sub-step of determination, for each column i:

• the contribution of the column to the total energy induced by the event in the set of photodetectors,
• the contribution of the column to the center of gravity, in X, of the event,
• the contribution of the column to the center of gravity, in Y, of the event.

b ₂ ) a substep of determination:

• the total energy induced by the event in all the photodetectors,
• barycentric coordinates (X ₀ , Y ₀ ) of the event with respect to the N photodetectors.

The method according to the invention then allows the exploitation of gamma-cameras with a high rate of counting, which is very advantageous in the case of measures of "attenuation correction by transmission "or" PET in coincidence ". high counting is achieved without restricting the number of photodetectors read. This is due to parallelism employed, and at the strong "pipelinage", that is to say at the succession of simple operations. The process according to the invention then makes it possible to accelerate the calculation of the center of digitized contributions from photodetectors, by parallelizing this calculation.

The same parallel processing can be applied to the determination of the barycentric coordinates (X ₁ , Y ₁ ) of the event and with the same advantages.

In the context of the invention, it is possible obviously replace the columns with lines, the principle of calculation remaining the same.

It is possible to proceed to a step preliminary detection of the presumed position of a event. We can then, around this position presumed to delimit a subset of N ' photodetectors out of the N photodetectors, only the signals from these N 'photodetectors being taken into account to perform steps b, c, d and e of the method according to the invention.

The invention applies in a particularly advantageous in the case where the photodetectors are photomultipliers a gamma camera. In particular, we can put in works the imaging techniques in correction transmission mitigation, and the techniques PET imaging in coincidence.

The invention also relates to a device for implementing the method described above.

Brief description of the figures

• la figure 1, déjà décrite, est une coupe schématique d'une tête de détection d'une caméra de type Anger connue ;
• la figure 2, déjà décrite, montre de façon schématique un dispositif de collection et de codage de signaux provenant de photomultiplicateurs de la tête de détection de la figure 1 ;
• la figure 3 représente l'interconnexion d'un ensemble de photodétecteurs,
• la figure 4 représente un dispositif de lecture d'un ensemble de photodétecteurs, pour la mise en oeuvre de l'invention,
• les figures 5A et 5B représentent des exemples de fonction F,
• la figure 6 représente la structure d'un système de calcul pour un mode de réalisation de l'invention,
• la figure 7 représente la structure d'un opérateur colonne pour un mode de réalisation de l'invention,
• la figure 8 représente la structure d'une partie d'un système de calcul pour un mode de réalisation de l'invention,
• la figure 9 représente une réalisation d'une autre partie d'un système de calcul pour un mode de réalisation de l'invention,
• la figure 10 représente un circuit associé à un photodétecteur, pour le traitement des données de ce photodétecteur ;
• les figures 11A et 11B représentent un signal analogique fourni par un photodétecteur (figure 10A), ainsi que le signal analogique numérisé correspondant (figure 10B) ;
• la figure 12 représente un mode de réalisation d'un dispositif pour la détermination d'une position présumée d'un événement.

In any case, the features and advantages of the invention will appear better in the light of the description which follows. This description relates to the exemplary embodiments, given for explanatory and nonlimiting purposes, with reference to the appended drawings in which:

• FIG. 1, already described, is a schematic section of a detection head of a known Anger-type camera;
• FIG. 2, already described, schematically shows a device for collecting and coding signals coming from photomultipliers of the detection head of FIG. 1;
• FIG. 3 represents the interconnection of a set of photodetectors,
• FIG. 4 represents a device for reading a set of photodetectors, for the implementation of the invention,
• FIGS. 5A and 5B show examples of function F,
• FIG. 6 represents the structure of a computing system for one embodiment of the invention,
• FIG. 7 represents the structure of a column operator for one embodiment of the invention,
• FIG. 8 represents the structure of a part of a computing system for one embodiment of the invention,
• FIG. 9 represents an embodiment of another part of a computing system for one embodiment of the invention,
• FIG. 10 represents a circuit associated with a photodetector, for processing the data of this photodetector;
• Figs. 11A and 11B show an analog signal provided by a photodetector (Fig. 10A), as well as the corresponding digitized analog signal (Fig. 10B);
• Fig. 12 shows an embodiment of a device for determining a presumed position of an event.

Detailed description of embodiments of the invention

The invention will be described in a manner detailed for photomultipliers of a gamma camera. However, this description also applies to any photodetectors, which do not necessarily part of a gamma camera.

FIG. 3 represents a set of photomultipliers 60, 60-1, 60-2, ... constituting a gamma-camera head. Each photomultiplier is identified by its position (i, j) in all the photomultipliers. More precisely, XC _{i, j} and YC _i denote the coordinates, along two axes X, Y, of the center of the photomultiplier i, j.

The signal from each of the photomultipliers is digitized and processed individually (integration, corrections, etc.). Each photomultiplier has a level of storage register that can memorize the contribution of each photomultiplier during the detection of an event. These aspects will be treated further in connection with Figures 10-11B.

The network of photomultipliers is organized in rows and columns, and all the outputs of the photomultiplier storage registers of a same column are connected on a bus 62 (bus column). Column buses can be collected in one 64 series bus.

An operator performing the various stages of the method according to the invention can be connected on the bus series. This operator then works in series, the time the longer the number of Photodetectors involved in the calculation is great.

It may also be planned to carry out a parallel processing: an operator performing the various stages of the process according to the invention is connected on the column buses 62. This gives a better counting rate.

When the photomultiplier network is not not too important, we can take into account all photomultipliers.

When the photomultiplier network is important (for example between 50 and 100 photomultipliers), a coarse position of the event is first determined, for example by a method to be described later, in connection with the figure 12.

Knowing the interaction zone of the event already allows to limit the number of photomultipliers to take into account to compute the position of the interaction. Indeed, only the two crowns of photomultipliers that surround the photomultiplier place of interaction carry in general information.

The accuracy of the presumed position being insufficient, we are often called to read a wider area than strictly necessary (for example, often up to 25 or 30 readings per event). The reading time of the registers of storage to handle an event is a passage obliged which directly influences the count rate of the machine (number of events handled by second).

In the case of parallel processing, means 66 can select each column independently of others. These means 66 are by example commanded by a sequencer (Figure 4).

Figure 4 illustrates more precisely a device for implementing the treatment in parallel.

A read sequencer 68 makes it possible to read the contents of the storage register of each photomultiplier. Let N _{i, j be} the contents of the storage register of the column photomultiplier i and line j. N _{i, j} is actually, for example, a digital integral of the signal delivered by the photomultiplier i, j in response to an event.

Means 70 make it possible to determine a presumed or gross position of the event. These means will be described further in a more in detail (Figure 12).

The command sequencer 68 the addressing of the columns by a multiplexer 74. The contribution of each column to the total energy, to the X component of the center of gravity (XC) and the component in Y of the center of gravity (YC) is transferred to a system of calculation 76, either by addressing the columns via the multiplexer, either directly by the sequencer of calculation 68.

The number of columns and rows taken into account around each presumed position is imposed by the inaccuracy of the determination. In Figure 4, 36 (6x6) photomultipliers are taken into account. In general, for a field of N photomultipliers it may be necessary to take into account N ₁ photomultipliers (N ₁ <N) as a function of the desired precision: in FIG. 4, N = 11 × 9 and N ₁ = ₆ × ₆ .

• les commandes nécessaires au multiplexeur 74 pour qu'il puisse orienter les bus colonnes à lire vers l'organe de calcul 76,
• les signaux de sélection de ligne de façon à présenter sur les bus colonne les registres de stockage commandés par la ligne n pour le premier temps de lecture, puis ceux commandés par la ligne n+1 pour le deuxième, et ceci jusqu'à ceux de la ligne n+5 pour le sixième temps de lecture,
• les coordonnées XC et YC des centres des photomultiplicateurs présentés sur les bus colonne.

Therefore, for each event corresponds an assumed position (PP) and, at each presumed position, a set of N ₁ (36) photomultipliers located on 6 rows and 6 successive columns is matched so that all photomultipliers that have stored information be retained. At each reading step (every 100 nsec), the read sequencer 68 provides, depending on the presumed position:

• the commands necessary for the multiplexer 74 so that it can orient the column buses to be read to the computing unit 76,
• the line selection signals so as to present on the column buses the storage registers controlled by the line n for the first reading time, then those controlled by the line n + 1 for the second, and this up to those of the line n + 5 for the sixth reading time,
• the XC and YC coordinates of the centers of the photomultipliers presented on the column buses.

The sequencer can be realized in form EPROM. At each presumed position corresponds a page memory in which the commands and values needed for calculation. This page is read online per line using a counter 72 which activates the low addresses of the EPROMs.

α) l'énergie de l'événement est la somme des contributions de tous les photomultiplicateurs qui entourent la position présumée :

Cette expression peut aussi s'écrire :

représente la somme des contributions des 6 photomultiplicateurs de la colonne i, l'énergie E étant la somme des énergies obtenues sur les 6 colonnes.

β) La position P₀ de l'événement est calculée par la méthode du barycentre :

• 
  où XC_i,j est la coordonnée en X du centre du photomultiplicateur situé sur la colonne i et la ligne j,
• 
  où YC_i,j est la coordonnée en Y du centre du photomultiplicateur situé sur la colonne i et la ligne j, Comme précédemment, on peut aussi écrire :
• 
  est la contribution au barycentre en X des 6 photomultiplicateurs de la colonne i,
• 
  est la contribution au barycentre en Y des 6 photomultiplicateurs de la colonne i.

The calculation of the energy and / or the raw position P ₀ , in X and Y, of an event can be summarized as follows:

α) the energy of the event is the sum of the contributions of all the photomultipliers that surround the presumed position:

This expression can also be written:

represents the sum of the contributions of the 6 photomultipliers of the column i, the energy E being the sum of the energies obtained on the 6 columns.

β) The position P ₀ of the event is calculated by the centroid method:

• 
  where XC _{i, j} is the X coordinate of the center of the photomultiplier located on the column i and the line j,
• 
  where YC _{i, j} is the Y coordinate of the center of the photomultiplier located on the column i and the line j, As previously, one can also write:
• 
  is the X-centroid contribution of the 6 photomultipliers of column i,
• 
  is the Y-centroid contribution of the 6 photomultipliers of column i.

We call the gross barycenter P ₀ the result obtained using steps α) and β) described above, as opposed to the weighted barycentre.

Once the gross barycentre P ₀ (X ₀ , Y ₀ ) is known, we can calculate the distance d _{i, j} from the center of each photomultiplier to P ₀ and weight the value N _{i, j} by calculating a new N _{i, j} called N ' _{i, j} , such that N' _{i, j} = F (N _{i, j} ), where F is a function of d _{i, j} (weighting function).

This function F is determined so empirical and adapted to each type of photomultiplier and at every collection geometry light.

• est inférieure à 1, pour d_i,j faible (pour minimiser la contribution du photomultiplicateur qui reçoit l'événement quand celui-ci est proche de son centre, ce photomultiplicateur contribuant peu à l'information sur le barycentre),
• est supérieure à 1 quand d_i,j est de l'ordre de grandeur de la dimension du photomultiplicateur (pour renforcer la contribution des photomultiplicateurs de première couronne, c'est-à-dire des photomultiplicateurs les plus proches du photomultiplicateur qui reçoit l'événement),
• tend vers zéro, lorsque d_i,j devient grand (pour diminuer la contribution des photomultiplicateurs au fur et à mesure qu'ils sont plus loin du lieu de l'interaction, car le rapport signal/bruit de leur contribution devient de plus en plus mauvais).

Function F, usually:

• is less than 1, for d _{i, j} low (to minimize the contribution of the photomultiplier that receives the event when it is close to its center, this photomultiplier contributing little to the information on the center of gravity),
• is greater than 1 when d _{i, j} is of the order of magnitude of the size of the photomultiplier (to reinforce the contribution of the photomultipliers of the first ring, that is to say photomultipliers closest to the photomultiplier which receives the event),
• tends to zero, when d _{i, j} becomes large (to decrease the contribution of photomultipliers as they are farther away from the interaction location, as the signal-to-noise ratio of their contribution becomes more and more bad).

A first example of function F (d) is given in FIG. 5A. This first example corresponds to 75 mm square photomultipliers. Values of F, for particular values of d (with a pitch of 5 mm), are given in Table I below. d F d F 0 0.8 80 1.18 5 0.809 85 1.13 10 0.826 90 1 15 0.85 95 0.831 20 0.883 100 0.663 25 0.916 105 0.494 30 0.95 110 0.325 35 0.983 115 0.117 40 1,016 120 0.1 45 1.05 125 0,065 50 1,083 130 0,035 55 1,116 135 0,015 60 1.15 140 0 65 1,176 145 0 70 1,194 150 0 75 1.2 - -

A second example of function F (d) is given in FIG. 5B. This second example corresponds to hexagonal photomultipliers of 60 mm. Values of F for particular values of d (with a pitch of 5 mm) are given in Table II below. d F d F 0 0.8 65 0.863 5 0.813 70 0.744 10 0.856 75 0.619 15 0.906 80 0,475 20 0.963 85 0.319 25 1,006 90 0,181 30 1,056 95 0.081 35 1,106 100 0,025 40 1.15 105 0,013 45 1,169 110 0 50 1,144 115 0 55 1,075 120 0 60 0,975 - -

From the point of view of the realization, the computation of the weighted barycenter P ₁ is carried out in the same way as the gross barycenter P ₀ , after replacing N _{i, j} by N ' _{i, j} . We then follow the calculation of the gross centroid by a weighting operation with the function F.

A computing device or system 76 for the implementation of the process according to the invention can have three sets schematically represented in Figure 6 and designated respectively by the references 78, 80 and 82.

The structure of the device 78 for determining P ₀ will be described more precisely in connection with FIGS. 7 and 8.

FIG. 7 represents means 90 associated with each column and subsequently called operator column.

The latter receives, in addition to the values N _{i, j} photomultipliers of the column, the coordinates XC _{i, j} and YC _{i, j} of the centers of the corresponding photomultipliers, for example provided by the read sequencer 68. The X coordinates are not necessarily identical for the same column because they must take into account the actual position of the photomultiplier within the gamma-camera field. The same applies to the Y coordinates. In the example given, each column bus output is connected to the input of a column operator 90.

Each column operator 90 accomplishes three operations, preferably in parallel.

A first operation consists in calculating the contribution of the column to the energy. For example, after being initialized at the beginning of the sequence, an accumulator 92 sums the values N _{i, j} of the 6 photomultipliers of the column and stores the result in a register 94 (CSsch). The outputs of the 6 registers (RSEcol1 to RSEcol6) are grouped on a common BECOL bus.

A second operation is to calculate the contribution of the column to the centroid in X. By example, after being initialized at the beginning of sequence, a multiplier-accumulator 96 performs the sum of the contributions to the center of gravity in X of the 6 photomultipliers of the column, and stores the result in a register 98 (RSXcol). The outputs of 6 registers (RSXcol1 to RSXcol6) are grouped on one BXCOL common bus.

A third operation is to calculate the contribution of the column to the centroid in Y. By example, after being initialized, a second multiplier-accumulator 100 makes the sum of contributions to the center of gravity Y of the 6 photomultipliers of the column, and stores the result in a register 102 (RSYcol). The outputs of 6 registers (RSYcol1 to RSYcol6) are grouped on one BYCOL common bus.

At the end of the 6 reading times, the 6 column operators who did their job become available for a new reading since the results of the first reading are stored.

In parallel with these calculation operations, the values N _{i, j} of the photomultipliers and the coordinates XC _{i, j} and YC _{i, j} can be stored in a system 104 of the FIFO type so that they can be used later.

A grouping of contributions to energy and at the coordinates of the centroid in X and in Y is then performed. This grouping will be described in using figure 8 where references 90-1, ..., 90-6 designate 6 column operators of the type described above in connection with Figure 7.

An accumulator 106, after being initialized at the beginning of the sequence, powered by the BECOL bus, calculates the sum of the six RSEcol1 registers1 to RSEcol6 and stores the result in a register 108 (ENERGY). The content of this register represents the sum of contributions to the energy of 36 photomultipliers surrounding the presumed position, so the energy of the event.

A second accumulator 110, powered by the BXCOL bus, sums the six registers RSXcol1 to RSXcol6, and stores the result in a register 112 (RXN). The contents of this register represent

A third accumulator 114, powered by the BYCOL bus, calculates the sum of the six registers RSYcol1 to RSYcol6, and stores the result in a register 116 (RYN). The contents of this register represent

The RSEcol, RSXcol and RSYcol registers are then released so that they can be used by column operators, and accumulators are well again available to process a new event.

Finally, the coordinates X ₀ and Y ₀ of the interaction point P ₀ of the event are calculated by performing two divisions using a divider 118: X 0 = RXN / ENERGY and Y 0 = RYN / ENERGY and this in less than 6 read times to be able to release the storage registers 108, 112, 116. The integrated pipelined integrated dividers (of the RAYTHEON 3211 type, for example) are capable of largely assuming these performances and even make it possible to use only one housing, performing the two divisions successively.

We thus obtain the position P ₀ of the event whose coordinates X ₀ and Y ₀ are stored in a register 120 (RX ₀ and RY ₀ ). In parallel with the divisions, the energy can be pipelined from the register 108 to a register 122 so as to release the register 108 for the next event.

A simple embodiment of the device 80 (calculation weighted barycentre: figure 6) is given in figure 9.

• 126-1 : récupération de la première valeur N_i,j et des coordonnées du centre du photomultiplicateur correspondant XC_i,j et YC_i,j (première valeur écrite dans la FIFO, donc première valeur lue) et stockage dans un registre d'entrée 128 de l'opérateur de pondération. Parallèlement mise en mémoire 130, à l'entrée de l'opérateur, de X₀ et Y₀. Cette mémorisation n'est pas répétée pour les 5 acquisitions de N_i,j, XC_i,j et YC_i,j suivantes puisque P₀ reste le même.
• 126-2 : calcul de dX=|XC_i,j-X₀|, et parallèlement de dY=|YC_i,j-Y₀|, et mise en mémoire 130, 132, 134, de dX, dY, N_i,j, XC_i,j et YC_i,j. Il y a ensuite libération des registres servant à l'étape 126-1, qui peuvent alors recueillir les valeurs relatives au photomultiplicateur suivant. Celui-ci est traité comme le précédent et ainsi de suite jusqu'au sixième.
• 126-3 : calcul de (dX)²=dX*dX, et parallèlement de (dY)²=DY*DY, et mise en mémoire 136, 138, 140 de (dX)², (dY)², N_i,j, XC_i,j, et YC_i,j. Puis il y a libération des registres de sortie de l'étape 126-2,
• 126-4 : calcul de d²=(dX)²+(dY)² et stockage de d², N_i,j, XC_i,j et YC_i,j dans des registres 142, 144 ; puis, libération des registres de sortie de l'étape 126-3.
• 126-5 : adressage d'une EPROM 146 contenant la fonction F'=f'(d²), par le registre contenant d². Ceci évite d'avoir à extraire la racine carrée de d², sachant qu'il est facile d'obtenir F'=f'(d²) quand on connaít F=f(d). Il y a ensuite stockage de F', N_i,j, XC_i,j et YC_i,j dans . des registres 148, 150 et libération des registres de sortie de l'étape 126-4.
• 126-6 : calcul de N'_i,j=F'*N_i,j et stockage de N'_i,j, XC_i,j et YC_i,j dans des registres 152, 154 ; puis il y a libération des registres de sortie de l'étape 126-5.

The values of N _{i, j} , X C _{i, j} , Y C _{i, j} , which were involved in the calculation of the gross barycenter were stored in FIFO memories generally designated by reference 124 (the memory 104 has already been mentioned above). (Figure 7) in which N _{i, j} is stored). The calculation is organized according to the following steps 126-1, ..., 126-6, for each weighting operator working on a column:

• 126-1: recovery of the first value N _{i, j} and coordinates of the center of the corresponding photomultiplier XC _{i, j} and YC _{i, j} (first value written in the FIFO, therefore first value read) and storage in a register of 128 input of the weighting operator. In parallel memory 130, at the input of the operator, X ₀ and Y ₀ . This memorization is not repeated for the 5 acquisitions of N _{i, j} , X C _{i, j} and Y C _{i, j} following since P ₀ remains the same.
• 126-2: computation of dX = | XC _{i, j} -X ₀ |, and parallel of dY = | YC _{i, j} -Y ₀ |, and storage in memory 130, 132, 134, of dX, dY, N _{i , j} , XC _{i, j} and YC _{i, j} . Then, the registers for step 126-1 are released, which can then collect the values for the next photomultiplier. This one is treated as the precedent and so on until the sixth.
• 126-3: calculation of (dX) ² = dX * dX, and parallel of (dY) ² = DY * DY, and storing 136, 138, 140 of (dX) ² , (dY) ² , N _{i, j} , XC _{i, j} , and YC _{i, j} . Then there is release of the output registers of step 126-2,
• 126-4: calculating d ² = (dX) ² + (dY) ² and storing d ² , N _{i, j} , XC _{i, j} and YC _{i, j} in registers 142, 144; then, releasing the output registers of step 126-3.
• 126-5: addressing of an EPROM 146 containing the function F '= f' (d ² ), by the register containing d ² . This avoids having to extract the square root of d ² , knowing that it is easy to obtain F '= f' (d ² ) when F = f (d) is known. There is then storage of F ', N _{i, j} , XC _{i, j} and YC _{i, j} in. registers 148, 150 and releasing the output registers of step 126-4.
• 126-6: computation of N ' _{i, j} = F' * N _{i, j} and storage of N ' _{i, j} , XC _{i, j} and YC _{i, j} in registers 152, 154; then there is release of the output registers of step 126-5.

Thus cut and pipelined the calculation of N ' _{i, j} is easily achievable because each computation step is simple enough to be performed during the duration of a reading step (typically 100 nsec).

The third computing subsystem 82 (FIG. 6), for calculating the weighted barycentre X ₁ , Y ₁ , has an architecture of the type of that described above in conjunction with FIGS. 7 and 8.

The weighted center of gravity output gives the new coordinates X ₁ , Y ₁ of the position of the event. The value of energy can be increased parallel to the calculations; in the same register 88, the energy and the coordinates of the event are obtained at the final output.

It remains to describe how the signal from each photomultiplier is detected and processed, in particular how an alleged position of the event can be calculated.

Figure 10 shows the part of the device associated with a single photomultiplier 60. The photomultiplier 60 connected to a converter current-voltage 262. In response to a detected event by the photomultiplier, we obtain a signal on the output 264 of the current-voltage converter 262, by example of the type of the one shown in the figure 11A.

The graph of FIG. 11A indicates, on the ordinate, the amplitude of the signal corresponding to the pulse and, as abscissa, the time. The amplitude of the signal and the time are indicated in arbitrary scale. t ₀ denotes the start time of the pulse provided by the photodetector and t ₁ the moment when the pulse returns to almost zero after having passed through a maximum. By way of indication, the duration corresponding to the interval t ₁ -t ₀ is of the order of one microsecond, in the case of a photomultiplier of a gamma-camera coupled to an NaI crystal (Tℓ).

The analog signal present on the terminal of output 264 is directed to a converter analog-digital 266. The latter samples each pulse of the signal into a number n samples, as shown in Figure 10B. Two consecutive samples are separated by a step, or clock interval p (the clock running at 1 / p Hz).

For example, the converter samples each pulse of the signal in n = 10 samples. For a signal of 1 microsecond, a sampling is then performed every 100 nanoseconds.

The analog-digital converter 266 is, preferably, a fast converter, of type "Flash" capable of operating at a frequency of the order of 10 to 20 megahertz.

The digital signal from the converter analog-digital 266 directed to an adder number 268. This summoner makes a sum of the samples sent to it by the analog-digital converter 266. The sum slippery is performed on a given number of samples. This predetermined number is equal, by example, at 10.

For each photodetector i, j, this sliding sum, or the digital integral of the signal associated with the event corresponds to the magnitude N _{i, j} already introduced above.

At the same time, the result of the summation carried out with the means 268 is stored in a register 271. The storage function can be composed of several registers to allow memorize several events temporally very relatives.

The value of the sliding sum is directed to means of comparison 270. The value of the sliding sum is compared with a threshold value predetermined set at an input 272 of the comparator 270. This comparator transmits on an output 274 a signal binary, representative of the result of the comparison (for example, 0 if the value of the sliding sum is below the set reference value and 1 if the value of the sliding sum is greater than the value reference).

In order to limit the duration of this overrun, it will only be validated during a window temporal centered on the maximum of the sum slippery. This allows to separate events close in time but geographically distinct on the detector field.

This window is positioned taking as reference the time of passage of the coded signal by a maximum. This detection is performed by the means 288 by comparing the current value of the output of the encoder to the previous value. When the current value is lower than the previous value, the comparator 288 emits a pulse. This pulse is sent to a shift register 290 whose delay n _{1 is set} to generate a time window centered on the maximum of the sliding sum. To take into account the inaccuracy (more or less a sampling step) of the determination of the position of the maximum of the coded signal, the time window will be activated during n ₀ no sampling with n ₀ ≥3 (for example n ₀ = 3), this choice of a minimum of three guaranteeing a minimum of simultaneity of the threshold crossing signals between the photomultipliers activated by the same event.

An AND gate 292, whose inputs are the signal obtained at the output of the comparator 270, and the output signal from the shift register 290, allows to obtain, on its output 294, a passing signal threshold at the desired moment in relation to the passage through the maximum of the digital signal.

FIG. 12 represents a device, compliant to the invention, for the processing of signals from several photodetectors 60, 60-1, 60-2. On this figure, references identical to those of the figure 10 designate similar elements or correspondents.

In this figure, we see that it is possible to take off at the output of the current-voltage converter 262, an analog signal 300, the type of which is described above in connection with Figure 11A. On the FIG. 12, the reference 302 designates globally all the analog signals taken from the other current-voltage converters 262-1, 262-2, .... All these signals fit into one analog summator 298 which delivers a signal S, sum of all the analog signals provided by a certain number of photodetectors, for example by all photodetectors. A device 304 makes it possible to deliver a pulse I during the passage of the signal S by its maximum. This device 304 comprises, for example, a differentiator (capacitance, amplifier and resistors between the input and the output of the amplifier); the output of this differentiator feeds a comparator which makes it possible to detect the passage to 0 of the output of the differentiator. Pulse I feeds the input of a shift register 306 whose pitch p is set by the clock H. The output 307 of this register is called memorization pulse and allows, in particular, to trigger the memory register 271 corresponding to the photodetector 60. It triggers also each memory register associated with photodetector. The delay of the shift register 306 is set so that the rising edge of the memory signal 307 is synchronous with the moment where we have to memorize the sums in the registers 271.

The set of photodetectors 60, 60-1, 60-2, ... is distributed for example in a network two-dimensional.

To identify the presumed position (or coarse) of an event in relation to this network two-dimensional photodetectors, we associate advantageously a memory access read at a first tracking direction in the network of photodetectors and a read access memory to a second tracking direction in the network of photodetectors. If we find this network by lines and columns, so we can associate a read access memory to locate a coordinate "Line" and read access memory to locate a "column" coordinate.

• un circuit OU (402) regroupe les sorties de type 294 des photodétecteurs d'une même colonne et génère un signal 422 actif lorsqu'au moins une de ses entrées est active. Il y a autant de circuits de type 402 que de colonnes,
• un circuit OU (412) regroupe les sorties de type 294 des photodétecteurs d'une même ligne et génère un signal 432 actif lorsqu'au moins une de ses entrées est active. Il y a autant de circuits de type 412 que de lignes.

More specifically, in a device according to the invention, of the type illustrated in FIG. 12, the outputs 294 which, when active, represent the photodetectors of the center of the interaction, are used in the following manner:

• an OR circuit (402) groups the type 294 outputs of the photodetectors of the same column and generates an active signal 422 when at least one of its inputs is active. There are as many type 402 circuits as there are columns
• an OR circuit (412) groups the type 294 outputs of the photodetectors of the same line and generates an active signal 432 when at least one of its inputs is active. There are as many 412 circuits as there are lines.

The type 422 signals are the addresses of a PROM 276 which is programmed to provide the coordinate (280) of the presumed position to the columns. Similarly, type 432 signals are the addresses of a second PROM 277 which is programmed from to provide the coordinate (281) of the position presumed in relation to the lines. The presumed position, represented by the pair of values (280, 281), is stored in a register 322, at the same time as memorize the contribution of all photodetectors in their respective registers (271). This memorization is triggered by the signal 307 generated by register 306.

Claims (24)

1. Process for determining the position (P₁) of an event in a coordinate system (X, Y) by means of a set of N photodetectors arranged in rows and columns, a random detector of a column of rank i and a row of rank j, being locatable in the coordinate system (X, Y), the event inducing a signal in the N photodetectors, the process involving the following steps:

   a) digitizing the signal supplied by each photodetector and calculating a value N_i,j representative of the energy supplied by each photodetector,

   b) calculating, as a function of N_i,j and the position of the photodetectors, an uncorrected position P₀ (X₀, Y₀) of the event relative to the set of photodetectors,

   c) determination of the distance D_i.j of each of the photodetectors relative to the position P₀,
   the process being characterized in that it also comprises

   d) the calculation of a corrected value N'_i,j = F'(N_i,j), in which F' is a function:

   which reduces (respectively does not modify) the value N_i,j for the photodetector corresponding to P₀ and for a certain number N₁ of photodetectors around P₀,

   which increases or does not modify (respectively increases) the value N_i,j for a certain number of photodetectors N₂ around the preceding N_i photodetectors,

   which tends towards 0 at higher values,

   e) calculating the position P_i of the event as a function of the position of the photodetectors and N'_i,j.
2. Process according to claim 1, in which the uncorrected position P₀ of the event is implemented by calculating the barycentric coordinates (X₀, Y₀) of said event as a function of N_i,j and the position (XC_i,j, YC_i,j) of each photodetector.
3. Process according to claim 1 or 2, in which the position P₁ of the event is implemented by calculating the barycentric coordinates (X₁, Y₁) of the event as a function of N'_i,j and the position (XC_i,j, YC_i,j) of each photodetector.
4. Process according to one of claims 1 to 3, comprising the following step before step e), for each photodetector at the edge of the field of N photodetectors:

   e₀) modify the value of the position (XC_i,j, YC_i,j) the new pair of values (XC'_i,j, YC'_i,j) being such that either the |XC'_i,j| > |XC_i,j| relation or the |YC'_i,j| > |YC_i,j| relation is satisfied, or both of these relations are satisfied.
5. Process according to one of claims 2 and 3, the barycentric calculations being carried out in sequence.
6. Process according to claim 2, in which the barycentric coordinates (X₀, Y₀) of the event could be determined using the following sub-steps:

   b₁) a sub-step in which the following are determined for each column i:

   the contribution of the column to the total energy induced by the event in the set of photodetectors,

   the contribution of the column to the X value of the barycentre of the event,

   the contribution of the column to the Y value of the barycentre of the event,

   b₂) a sub-step in which the following are determined:

   the total energy induced by the event in the set of photo-detectors,

   the barycentric coordinates (X₀, Y₀) of the event with respect to the N photo-detectors.
7. Process according to claim 3, in which the barycentric coordinates (X₁, Y₁) of the event could be determined using the following sub-steps:

   e₁) a sub-step in which the following are determined for each column i:

   the contribution of the column to the total energy induced by the event in the set of photodetectors,

   the contribution of the column to the X value of the barycentre of the event,

   the contribution of the column to the Y value of the barycentre of the event,

   e₂) a sub-step in which the following are determined:

   the total energy induced by the event in the set of photodetectors,

   the barycentric coordinates (X₁, Y₁) of the event with respect to the N photodetectors.
8. Process according to one of claims 1 to 7, also comprising an additional step prior to step b):

   a'₁) to determine the presumed position of an event.
9. Process according to claim 8, also comprising a step:

   a'₂) to delimit a subset of N' photodetectors among the set of N photodetectors around the presumed position of the event, only the signals from the N' photodetectors in this subset being processed according to steps b, c, d and e.
10. Process according to one of claims 1 to 9, the photodetectors being photomultipliers of a gamma-camera.
11. Imagery process in correction of transmission attenuation, embodying the process according to claim 10.
12. PET coincidence imagery process embodying the process according to claim 10.
13. Device for determining the position of an event in a coordinate system X, Y, particularly with respect to set of N photodetectors, comprising:

    a) means (266, 268) of digitizing signals output by each photodetector, and of calculating a value N_i,j representing the energy of the signal output by each photodetector,

    b) means (76, 78) of calculating an uncorrected position P₀ of the event, characterized in that it also comprises:

    c) means (80, 130, 132) of determining a distance d_i,j of each photodetector relative to the uncorrected position,.

    d) means (80, 146, 152) for calculating a corrected value N'_i,j of the energy supplied by each photodetector, said means (80, 146, 152) performing the said calculation in accordance with the formula N'_i,j = F'(N_i,j), in which F' is a function:

    which reduces (respectively does not modify) the value N_i,j for the photodetector corresponding to P₀ and for a certain number N_i of photodetectors around P₀,

    which increases or which does not modify (respectively increases) the value N_i,j for a certain number of photodetectors N₂ around said preceding N₁ photodetectors,

    - which tends towards 0 for higher values,

    e) means (82) for calculating the position P₁ of the event, said means being controlled by the means for calculating the corrected values N'_i,j = F' (N_i,j).
14. Device according to claim 13, the means of calculating the uncorrected position P₀ of the event being means of calculating the barycentric coordinates of the said event.
15. Device according to claim 13 or 14, the means of calculating the new position P₁ of the event, being the means for implementing a barycentric calculation of this new position.
16. Device according to claim 13, also comprising means:

    d') of modifying the value of the position (XC_i,j, YC_i,j) of the photodetectors located at the edge of the photodetectors field, to a position (XC'i,j, YCi,j) in which the [XC'i.j] > ]XCi,j] relation or the ]YC'i,j] > [YCi,j] relation is satisfied, or both of these relations are satisfied.
17. Device according to one of claims 13 to 16, also comprising:

    a set of N photodetectors (60, 60-1, ..., 60-2) laid out in rows and columns,

    means of grouping the photodetectors in the same column on a column bus (62),

    means of grouping the column buses on a serial bus (64).
18. Device according to claim 17, an operator including all means b) to e) being connected to the serial bus and performing serial processing of the data.
19. Device according to claim 17, an operator grouping means b) to e) being connected to the serial bus and implementing a parallel data processing.
20. Device according to claim 13, comprising means (90, 90-1, ..., 90-6) of determining the following for each column i:

    the contribution of the column to the total energy induced by the event in the set of photodetectors,

    the contribution of the column to the X value of the barycentre of the event,

    the contribution of the column to the Y value of the barycentre of the event,
    and means (106, 110, 114) in which the following are determined:

    the total energy induced by the event in the set of photodetectors,

    the barycentric coordinates (Xo, Yo) of the event with respect to the N photodetectors.
21. Device according to claim 15, also comprising means for:

    b') determining the following for each column i making use of the values of the weighted signal N'i,j:

    the contribution of the column to the total energy induced by the event in the set of photodetectors.

    the contribution of the column to the X value of the barycentre of the event,

    the contribution of the column to the Y value of the barycentre of the event,

    c') determining the following making use of the values of the contributions obtained in b')

    - the total energy induced by the event in the set of photodetectors

    - the new barycentric coordinates (X1, Y1) of the event with respect to the N photodetectors.
22. Device according to any one of claims 13 to 21, also comprising means (70) of detecting the presumed position of an event.
23. Device according to claim 22, also comprising means of delimiting a subset of N₁ photodetectors among the set of N photodetectors around the presumed position of the event.
24. Device according to one of claims 13 to 23, the photodetectors being the photomultipliers in a gamma-camera.

Images